The tone of French intellectual life in the last decade of the Second Empire and first years of the Third Republic was set by a scientism that labeled itself positivist. The nation's elite students dove into the world eager with the self-assurance of the newly converted, having found in a single idea the key that would turn all locks. Their idée-clé was a vague vision of science composed of mechanical causality and evolutionary biology; the doors it opened led in diverse directions, from explanations of miracles to etiologies of diseases. Academic critics of the scientistic world view hunkered down in their redoubts, fearful that they were fighting a losing battle against the future. By the middle of the 1870s, however, the limits of French positivism had begun to show and some doors, it seemed, would remain locked. Pushing against the positivism of his peers, Emile Boutroux, in his 1874 thesis De la Contingence des lois de la nature, presented an argument that both changed the terms of debate in his field and resonated with scholars for generations to come. In a typical irony of academic life, however, Boutroux's work was hailed not primarily for its originality, but for its limpid summation of all that had come before. Two generations would pass before the central insight of De la Contingence had run its course and in the meantime Boutroux's role in forging a new relationship between philosophy and science would be partially understood and insufficiently appreciated, even by Boutroux himself. With the benefit of hindsight, however, it is clear that Boutroux gracefully bridged two salient divides of the early Third Republic; his ability to make science palatable to a segment of the Catholic population has been addressed elsewhere, his influence on republican ideology has not.Footnote ¹

The insight at the heart of Boutroux's thesis was the recognition of the fact that science had outpaced philosophical innovation, that philosophers could no longer meaningfully claim that theirs was the queen of the disciplines or that they had any unique purchase on a realm of metaphysical fact beyond the ken of scientific investigation. Boutroux was well positioned to recognize the power of modern science; he was related by marriage to the great mathematician Henri Poincaré and tied by bonds of friendship to a number of physicists, mathematicians, and astronomers. In recognizing the emerging primacy of science, Boutroux reshuffled the terms of a debate over free will and determinism that had been key to French political and religious self-understandings in the mid-nineteenth century. The impetus behind Boutroux's groundbreaking thesis was not only philosophical, but also, in part, religious. Supporting a traditional philosophical argument for free will on a novel scientific foundation gave Boutroux's Catholic beliefs a broader worldly platform than did faith alone. Boutroux's suggestion that contemporary scientific theories were compatible with the free will that was central to his faith, as well as his ability to guide theological interests into scientific and even republican channels, make him an intriguing anomaly of his era.

The competition between religion and science over the right to speak the truth is always at most a step away from barging into the political realm. Liberal political theorists, in particular, still struggle to defend the epistemological privilege of science in public discussion and in education without violating the procedural neutrality that characterizes liberal institutions. The political stakes of debates between science and religion were higher in the nineteenth century than they are today. Political thinkers of all stripes were drawn to the dilemma posed by the expansion of suffrage to populations that were still largely Catholic and, many feared, hostile to democratic government; for republicans, the solution was to replace religious with scientific education and, thus, faith with reason. By the turn of the century, the conflict between antidemocratic religious movements and supporters of the secular state had reshuffled political alignments, changed the social position of religious figures, and led a significant reorientation of the social sciences.Footnote ² The political impact of the fin de siècle realignment is well known, and the obituary of mid-century liberal dreams has been told often and well. Two decades before the rise of Austria's Christian-Socials and French anti-Dreyfusards occasioned a crisis of confidence in rational politics, however, another debate raised questions about the status and nature of reason. The more limited context of the 1870s, before religious antirepublicanism fused with secular anti-Semitism, provides a focused case study in the philosophical dimension of the meetings of Catholicism and science. The debates of the 1870s were fought at great distance, between believers in an interventionist God and positivist visions of a mechanically determined universe, but the ground between revelation and reason was not entirely barren. Boutroux's attempt to reconcile his faith with his esteem for mathematics and physics lay between the lines of science and faith. Reason, according to Boutroux, was necessary to understand the world, but it was not the code of pure intelligibility envisioned by positivists. Rather, it was the guarantee that free will would not devolve into caprice.

Boutroux's religion served as a distorting mirror to nineteenth-century observers. On one hand, he was well received in Catholic circles for the unambiguously Christian inspiration of his work. On the other hand, his students tended to minimize the role that religion played in Boutroux's philosophy. His students were republicans, often Jews; writing at a fin de siècle in which the credibility of Catholicism was at a nadir with the French left, they had little reason to emphasize the religious roots of their scholarship. Religion did, however, play a significant role in Boutroux's thought, though not in ways that the traditional historiography of the Third Republic has conditioned us to expect. The theological concerns driving Boutroux's novel philosophical take on modern science fly in the face of an interpretation of the early Third Republic, until recently dominant, which posits an impregnable division between secular republicanism and Catholic reaction.Footnote ³ For a century, the heroic phase of the Third Republic was viewed through the lens of laïcité and fiery anticlerical rhetoric was taken at face value as a sign of intractable republican opposition to religion. Historians have begun to expose the weaknesses of Manichean descriptions of the republic by showing the role that Protestantism played in republican and liberal ideologies of the 1860s and 1870s. Searching for a more nuanced understanding of 1860s religious disputes, Sudhir Hazareesingh has rightly warned that the historiographic temptation to draw “the moral and political contours of republican doctrine around the conflict between Church and State often reduces the ideological history of the period to a bipolar clash between rationalism and religion.”Footnote ⁴ In a different register, Venita Datta has shown the overlapping concerns of the Dreyfusard and anti-Dreyfusard literary avant-garde, breaking down what she calls a historiography of “monolithic blocs,” and Robert Gildea has stressed the accommodationist practice that accompanied the swingeing rhetoric of republican anticlericalism.Footnote ⁵ Recent historiographic nuance on the role of Protestantism notwithstanding, little evidence of liberal Catholicism has emerged. Even the most recent work on French philosophy maintains a strict division between Catholic philosophers, who supposedly saw philosophy as unchanging in the face of scientific advance, and secular republicans, who saw the mind itself as evolving over time.Footnote ⁶ Examining Emile Boutroux's early work allows us to expand our understanding of religion and republicanism, bringing another Catholic into the republican fold.

Though the history of science and the history of political ideologies are rarely placed in dialogue, they mirror each other, with classic nineteenth-century accounts of science echoing the hopes and fears of the wider polity. Contemporary writers imagined a republic that would lean on positive science to displace religion as the core of French education. Scientific knowledge would be the glue binding society together. Yet the anticlerical self-representation of French science is belied by the fact that some of the most articulate accounts of mathematics and physics, notably those of Paul Tannery, Edouard Le Roy, and Pierre Duhem, were motivated by religious and ethical commitments. These commitments were constructed at times against antirational currents in Catholicism and at times against the extremes of positivist scientism and materialism. Boutroux, a pioneer among antipositivist philosophers of science, described his project in limpid terms:

Following the lead of political historians, who have rendered untenable the vision of republican political thought as implacably atheist, I argue that we cannot understand the rationalist philosophy of the late nineteenth century (which, increasingly, took science as its object) as innocent of religious concerns. The antagonisms between religion and positivist science that seem so absolute at a distance could play out within individual minds as productive tensions, motivating bursts of creativity and fascinating attempts to hold together psychological commitments that, when articulated in the political sphere, sounded increasingly incompatible.

* * *

The debates to which Boutroux's thesis responded would have been readily apparent to French scholars of the 1870s. A pervasive concern with free will and mechanical determinism permeated academic discussion and occasioned the publication of several books a year either defending the irreducibility of human conscience or articulating schemes for reducing thought to material and biological determinants. This debate, so pervasive in Boutroux's day that it need not have been stated, is no longer transparent, so it is worth laying out the primary context in which De la Contingence could or would have been read.

The texture of French thought of the early nineteenth century has become increasingly clear in recent years as historians have drawn attention to the thread of liberalism running from the post-Revolutionary decades to the end of the century, and in a different vein, rediscovered the strange era of Victor Cousin's intellectual and institutional dominance in the 1830s.Footnote ⁸ Cousin, who had begun his career as a suppléant to philosopher Pierre-Paul Royer-Collard, rose through the academic administration to the pinnacle of minister of public instruction under the July Monarchy. From his lofty perch, Cousin could determine the philosophical curriculum across France, personally vet professorial hires, and judge philosophy dissertations throughout the country.Footnote ⁹ The great power that Cousin wielded was out of all proportion to the rigor and originality of his thought. Cousin held that instead of aiming to create total systems, philosophers should patch together congenial world views out of bits and pieces taken from other thinkers (emphasis on the “eclectic” in eclectic spiritualism). Philosophical borrowings were sewn together with the goals of establishing a stable and autonomous vision of selfhood and integrating all knowledge into a single system. Cousin left space for the Church to reign on issues of morality and the regime in power to rule on issues of politics. This malleable philosophy won adherents in the government, which in turn eased Cousin's climb through the university administration.Footnote ¹⁰

Cousin dominated the academy of the July Monarchy, but by Boutroux's student years Cousin had fallen from power, and with his fall went the self-assurance of philosophical disciplinary primacy. The central question of Boutroux's thesis emerged from the breakdown of Cousin's thought. The 1860s had witnessed the growth both of a positivist insurgency, which opposed the Empire and the regnant intellectual orthodoxy, and of a nascent neo-Kantianism that rejected positivism while seeking a level of analytical rigor never found in eclectic spiritualism. An anemic portrait of the German philosopher emerged from Cousin's teachings, and though translations appeared sporadically in the 1840s, they were the work of isolated scholars outside the academic mainstream.Footnote ¹¹ Only when Boutroux's mentor, Jules Lachelier, discovered Kant and began to send students, Boutroux included, to Germany did France seriously engage with critical philosophy. Though his own 1869 trip to Germany was cut short by war, Boutroux was profoundly marked by the lectures of neo-Kantian Eduard Zeller. Boutroux was so taken with Zeller's work that Lachelier professed himself unable to distinguish Boutroux's own thoughts from his reports on Zeller's ideas.Footnote ¹² Zeller had taught Boutroux of Kant's desire to found philosophy on contemporary scientific knowledge and of his search for the limits of scientific discovery; Boutroux took both of these Kantian hallmarks over as his own in his first major project.Footnote ¹³ In the twentieth century, Boutroux would be widely acknowledged as a founding member of French neo-Kantianism, but he should not be confused with a strict Kantian.Footnote ¹⁴ As will become apparent, his writings take an idiosyncratic interpretation of Kant as their starting point, and his conclusions are often at odds with Kant. Boutroux would argue in his thesis that underneath the appearance of uniform, lawful causality, there exists a radical contingency, which was an effort to bring Kant's noumenal realm forward into the world of perception while changing the content of that noumenal realm (where contingency, rather than causality, now reigned.) In this, at least, he was inspired by Kant, though he departed from the German philosopher with his arguments that theory and practice are not necessarily coherent, his implication that time must exist outside the structures of our understanding, and his denigration of causality.Footnote ¹⁵ Despite his questionable readings of Kant, I will follow Boutroux's contemporaries in referring to his work as neo-Kantian, since the obvious alternatives obfuscate more than they clarify.

The struggle between the crumbling orthodoxy of Cousin's thought and the emergent positivist and neo-Kantian movements was played out in the debate over free will and determinism that provided the most overdetermined philosophical trope of the 1860s and early 1870s. The debate was a metaphor for the struggles between republicans and the Church, and a proxy for the disciplinary battle between the sciences and the humanities. Positivists attacked the Cousinian vision of an autonomous selfhood beyond the reach of scientific investigation, taking eclectic spiritualism as the philosophical face of the Church and the Empire. Cousin's students responded with variety of arguments defending the autonomy and indeterminacy of intellectual life. Both sides, however, agreed on a basic understanding of the material universe, one informed by Newtonian mechanics and Euclidean geometry in which the world was made up of endless chains of cause and effect. Boutroux's thesis would provide an alternative to the stale dispute between science and philosophy over whether or not human thoughts and actions were governed by material causes.

Boutroux facilitated the search for the proper context in which to read De la Contingence with his gift of “La Philosophie en France depuis 1867,” a report presented in 1908 to the third meeting of the International Congress of Philosophy.Footnote ¹⁶ Taking the baton passed by his teacher, Félix Ravaisson, who had, in 1867, overseen the reckoning of two-thirds of a century of philosophical thought, Boutroux produced a report tilted toward the analysis of his contemporaries and students.Footnote ¹⁷ His discussion of the successive fractures in French philosophy, which produced psychology, sociology, and a nascent political science, is fascinating in itself, but more pertinent are his opening remarks on the state of philosophy in his student years, the 1860s and early 1870s.

Among the first thinkers signaled by Boutroux as still relevant in the years of his youth, Etienne Vacherot stands out for his influence. In his 1858 La Métaphysique et la science, Vacherot displayed a confidence in the reign of philosophy over the terrain of the mind that would seem dated within a decade. Vacherot, who had succeeded Cousin at the Sorbonne, saw philosophy and the natural sciences as largely discrete disciplines with well-delimited areas of investigation, natural science confining itself to practical problems of the empirical world, while philosophy addressed the entirety of mental life.Footnote ¹⁸ The very form of Vacherot's La Métaphysique et la science, written as a dialogue between the metaphysician and the scientist, exhibited the epistemological autonomy of the two positions, each figured as a discrete, rational, self-sufficient interlocutor. Continuing the eclectic spiritualism of his mentor, Vacherot was hostile to German thought and comfortable with a division of labor that left the entire physical world to the natural sciences, allowing metaphysics to explore the realm of imagination and the soul.Footnote ¹⁹ Vacherot's enormous confidence in the power and relevance of philosophy was broadly stated, but hardly exceptional for its time. Ten years after Vacherot's publication of La Métaphysique et la science, another one of Cousin's students and an examiner on Boutroux's thesis, Elme Caro, would express unease with the solidity of the spiritualist edifice so recently taken for granted.

In his 1868 Matérialisme et la science, Caro had to take the claims of science, and even of positivism, much more seriously than had Vacherot, arguing that metaphysics was faced with a choice between extinction and an “honorable peace” with the sciences.Footnote ²⁰ For Caro, such a peace was contingent upon the victory of a scientific vision that held a place for philosophical knowledge. Distinguishing between an “experimental school” of science (exemplified by, and for all appearances exhausted in, Claude Bernard's Introduction à l’étude de la médecine expérimentale) and the rigorous positivism of Comte's epigones, Caro argued that scientists and philosophers of the experimental school left room for metaphysics and religion, while no negotiation was possible with positivism. What interests me in Caro's argument is not its self-serving typology of scientists, divided into those who allow for metaphysics and those who do not, but rather the fundamental assumptions that it makes, assumptions that mirror those of Caro's positivist opponents. Between the two camps there was a wide, though unacknowledged, agreement about the nature of scientific practice and the relationship of science and freedom. Caro, one of the foremost critics of positivism, cedes the ground of the material and biological world at the outset, defining science as causal and determined: “The essential character of all scientific facts is that they are determined, or at least determinable. To determine a fact means to attach it to its immediate cause and to explain it by this cause,” and further, “one must admit as axiomatic that in identical conditions, all phenomena are identical, and that as soon as conditions cease to be the same, the phenomena cease to be identical.”Footnote ²¹ The emphasis on scientific determination ruled by rigorous chains of causality was widely shared by Caro's spiritualist allies and positivist opponents alike. What typified the antipositivist position was the defense of metaphysical terrain; Caro draws the line at intellection, arguing that positivism had yet to show that (physical) brains think.Footnote ²² Lacking any causal body–mind link, positivism could not hope to venture into the philosophical world. Caro's argument is a defense of metaphysics, conceived as the study of that portion of thought unreachable through physical or biological investigation, but his positivist opponents tended to see it as a defense of religion, and not without reason. A list of recent eclectic spiritualist polemics against positivism compiled by Caro's colleague Paul Janet in 1865 comprises the following titles: Essay on Religious Philosophy, Natural Religion, Moral Studies on the Present Day, and The Idea of God and Its New Critics.Footnote ²³ Whether they were religious or secular, however, spiritualist works shared Caro's defense of mind–body dualism, which typified, mutatis mutandis, all of the antipositivist tracts of the 1860s and early 1870s.

Looking back at the intellectual field of his youth with thirty years of hindsight, Boutroux portrayed a double break with Cousin's philosophy as the main event of the 1860s; this break was executed in one direction by positivists inspired by the recently translated Darwin and Spencer, and in the other by the French discovery of Kant. Yet tellingly, both variants of the antispiritualist position display an understanding of the relationship between causality, science, and freedom similar to that articulated by Vacherot and Caro. What Boutroux would later call the neo-Kantian position was really an evolution of Cousinian thought that only became avowedly Kantian with Boutroux and only became a defensible interpretation of Kant with his students, but in Boutroux's eyes it included himself and his mentors, the two philosophers whom he credits with the reform and renewal of post-eclectic French philosophy: Félix Ravaisson and Jules Lachelier. Addressing the free-will debate, Ravaisson and Lachelier had both claimed that the human capacity for intellection separates us from all other beings. Lachelier had argued that because all of the accessible (phenomenal) world is reliant upon human intelligence, humans maintain a spiritual autonomy denied to the rest of the universe. Metaphysics remained, however, for both Lachelier and Ravaisson, free from the laws of scientific necessity.

At the wellspring of French neo-Kantianism, one of Boutroux's peers, Alfred Fouillée, had written a dissertation in 1872 that began from the self-reflective discoveries of Cousinian psychology, finding freedom in the conceptual possibilities that precede action.Footnote ²⁴ Fouillée argued that the idea of liberty alone was enough to guarantee its reality because ideas have real-world impact (the idées-forces for which Fouillée was known in his day) and thus the capacity to undermine predetermined physical interactions. Like Boutroux, Fouillée rejected the binary opposition between freedom and necessity (a position he mocks when he finds it in the work of some eclectic spiritualists) in order to show that liberty emerges from our everyday understanding of determinism. While the conceptual framework of Fouillée's thesis is in several ways more convincing than Boutroux's, his argument is predicated upon an anthropology, making it a tenuous reply to positivism, which denied the human exceptionalism implicit in Fouillée, as well as in Caro and other spiritualists. In the decade and a half before Boutroux, then, the disparate attempts to save free will by limiting the scope of scientific investigation were unified less by the structure of their arguments than by what they took for granted. None of the major critics of positivism questioned the mechanistic vision of the universe, and all tried to exempt humans from the unbreakable chains of cause and effect that seemed to govern the material world.Footnote ²⁵

On the other side of the struggle to replace the disintegrating Cousinian orthodoxy lay positivism. French positivism was something of a sensation in student circles of the late Empire, though it never gained a firm foothold within philosophy departments.Footnote ²⁶ Auguste Comte, the movement's founder, had been pushed to the margins of the academy by Cousin, and after his death his disciples had split over the sometimes bizarre religious elements of his philosophy, giving the school an unpropitious adolescence. By the 1860s, however, the popularity of Comte's thought was again waxing. In the late Empire positivism developed a protean visage: at its simplest, it was the claim that an author intended to stick to the facts and eschew metaphysical or religious speculation. In a more restricted sense, most French positivists of this era saw the natural sciences as the template for all knowledge and all useful belief. At the most specific, some positivists, though far from all, claimed to follow to the letter the methods laid down by Comte.

Positivism was given a huge boost in the Second Empire by its status as a political pariah. As students, in particular, turned toward republicanism, the Empire's censure of positivist scholarship gave it the cachet that comes with the whiff of rebellion. With the sanction of Comte's disciples on both sides of the school's schism—Emile Littré and Pierre Laffitte were both republicans—students ignored Comte's political conservatism and treated positivism as the philosophical analogue of democratic politics.Footnote ²⁷ In addition, a slew of translations of Herbert Spencer's work in the 1860s showed how science could tackle not only the inert but also the biological world with startling results. Darwinian scholarship redoubled confidence in science and eroded the distinction between life and brute matter as they both came under the purview of the sciences. Because of the division within positivist ranks and the loose usage of terminology, positivism was often simply associated with scientism and began to lose much of its analytical value. In his own analysis of the philosophical field, Boutroux tellingly abandons positivism as a category, grouping scientifically minded philosophers under their emerging disciplines of psychology and sociology.

Positivism never took over philosophy departments, but Boutroux and his peers clearly reacted to the school's reductive view of human existence, an extension of the determinist vision of the material world so pervasive among metaphysicians. In its extreme form, the positivist reduction took the guise of materialism, witness Littré's definition of the human soul as “anatomically, the sum total of the functions of the brain and spinal cord: physiologically, the sum total of the functions of cerebral sensation.”Footnote ²⁸ Littré drew out the ethical implications of his materialism in his influential French dictionary, arguing in a definition of esprit that

Littré, a disciple of Comte, had a narrower audience than Hippolyte Taine, the dean of unofficial positivism, who applied scientific methods to the study of literature and history to great renown. Famous for his theory that national literature could be analyzed under the “scientific” rubrics of “race, moment, and milieu,” Taine, like Littré, rejected metaphysical ideas as “phantoms.” In his De l'Intelligence (1870), Taine reduced mental functions to their atomic components to show that biological states have direct intellectual and emotional effects. In his 1857 polemic against eclectic spiritualism, Taine had argued that all metaphysical conceptions, such as force, substance, and nature, were in essence different ways of talking about causality. For spiritualists, Taine argued, causes are “distinct beings . . . resembling the internal force that we call our will”; for positivists and scientists—the two were the same for Taine—discussion of first causes gives way to “the sum and direction of the chemical reactions and physical actions that compose life, and to ordering the experimental laws that sum up the facts observed in our universe.”Footnote ³⁰ The fundamental assumption shared by eclectic spiritualists, neo-Kantians, and positivists remains, that the entire observable world is law governed, materialist, and predetermined.

The context of French philosophy in the two decades from 1855 to 1874 was set by the precipitous disintegration of the eclectic-spiritualist orthodoxy. While several of Cousin's students tried to save their school, revisionist idealists who would eventually be known as neo-Kantians rapidly gained power within philosophy departments, while support for positivism grew in student ranks and in pockets of the academy. What appeal positivism had among philosophers waned quickly, but it retained its force in the new disciplines that split off from philosophy, in particular psychology and sociology. The disintegration of the eclectic-spiritualist orthodoxy could be described in sociological language as an institutional or disciplinary dispute,Footnote ³¹ but it was also an intellectual conflict, fought out on the terrain of the human mind. At stake was the question of whether humans have the capacity to act autonomously, or whether we are entirely determined by our biological being. From the positivist side, biological determinism implied predictability and therefore the potential for scientific investigation of the mental universe that could parallel the rapidly expanding exploration of the physical world. From the eclectic-spiritualist side, materially determined human consciousness would spell an end to ethical norms and the relevance of religion. Going a step further, neo-Kantian philosopher Charles Renouvier even argued that positivism posed a threat to democracy, since voting is only a meaningful act if people freely choose their actions.Footnote ³²

Beneath the surface froth of animosity between positivists, eclectic spiritualists, and neo-Kantians, a set of basic assumptions lay quietly unexamined. These assumptions acted as the doxa of the free-will debate.Footnote ³³ All interlocutors in the debate shared a single definition of science as the study of mechanically determined systems operating through chains of cause and effect. All sides agreed that there was a single dispute to be settled: did the material world include human beings, or did it not? Boutroux's exceptional move was breaking with this binary, and it is to this element of his thesis, De la Contingence des lois de la nature, that I now turn.

Like Fouillée's Liberté et déterminisme, Boutroux's thesis was an explicit attempt to answer the philosophical problem of his day: did free will exist and, if so, what was its relationship to scientific knowledge. Boutroux broke, however, with the intellectual doxa of the 1860s by accepting the analytical power of science, while contesting the claim—held across the spectrum of the free-will debate—that science was inherently determinist. The problem with positivism was not so much the supposition, contested by eclectic spiritualists, that scientific theories provided a universal template for knowledge, as it was the simplistic understanding of science that the school expounded. Where critics of positivism had accepted a mechanist vision of mathematics and physics, Boutroux took the critique of positivism to the heart of these flagship disciplines. Boutroux based his arguments in De la Contingence on the claim that philosophers had generally misunderstood how science actually works, even when it deals with inanimate objects. Positivists and eclectic spiritualists alike had believed that physical objects are fungible, that there can be identical objects reacting identically to stimuli. The fungibility of physical objects was the sine qua non of purely mechanistic visions of the universe, since natural laws had to apply equally to all like objects. For materialists, such mechanically determined relations actually were the world; for eclectic spiritualists, these relations governed the entire physical world, though not the world of intellection. This nearly universally held view, Boutroux argued, was wrong.

Boutroux began his thesis by asking whether contingency might not reside in the relations between things themselves, rather than solely in the insufficiencies of our ideas of the world, which would imply a failure of the determined vision of positivism.Footnote ³⁴ He attacked the mechanistic vision of the universe on three increasingly concrete (and increasingly convincing) levels: ontological, nominal, and systemic. At an ontological level, Boutroux argued that the phenomenal universe never exactly matches the universe of pure being, and so scientific ideas never quite fit the (impure) phenomena of the world.Footnote ³⁵ At the nominal level, Boutroux argued that no two real-world objects could ever be exactly alike, at a logical limit, because no two could ever occupy the same position in space and time, and thus never be subject to exactly the same constellation of forces. “Identity exists only in words,” he pointed out; “nature presents us only with resemblances, not with identities.”Footnote ³⁶ Minuscule differences between like objects usually presented no obstacle to prediction, but they were a chink in the armor of the mechanist argument.

From the foundation of minuscule variations in worldly objects, Boutroux built upwards to the level of systems, which he envisioned as collections of objects united by interactions taking place over time. A simple set of colliding balls might be a system, but so too would be the daily conversations that make up a university department. Positivists, according to Boutroux, thought that the entire world could be perfectly described in mathematical terms as one enormous closed system, comprising simple physical interactions. This meant that time existed for the positivist only as the playing out of a preordained series of causes and effects. Boutroux, on the other hand, argued that no theorem could describe reality perfectly. Because of the ever-present possibility of a chance event, a “contingency,” any chain of apparent causes and effects could veer off course. The possibility of chance occurrences led Boutroux to argue that physics had to include an element of history, that without studying the history of a physical system we could not rule out the possibility that its current form was the result of contingency rather than the theoretically predicted chain of causes and effects. Boutroux was wrong on this point, but his contributions to physics are much less interesting than his philosophical heritage. The historical aspect that Boutroux demanded of physics was one of the original moments in the oft-noted interest in history displayed by French philosophers of science.Footnote ³⁷

Against the positivists’ mechanistic description of the universe, Boutroux argued that the process of building categories out of the seemingly infinite number of unique objects contains inherent imperfections. Because all objects are unique, every category that groups together more than one object fails to be perfectly adequate to its contents. At a simple level, categories can describe the objects that they group together with near-perfect precision. However, Boutroux argued, scientific knowledge is grounded in increasingly dense layers of categories, so the more complex the scientific claim the more imperfect the categories on which it will be based. At the level of complex systems, understanding the unique history of the system becomes the only way to grasp it in its present form—in Boutroux's words, “the study of the history of beings acquires . . . a singular importance.”Footnote ³⁸ History could adjust for contingency in the development of an object or a system in a way that the a priori logic that typified both positivist and eclectic-spiritualist accounts of science could not. By following the historical development of a system, scientists could correct for minor variations in putatively like objects and thus account for contingency. Although Boutroux's intention was to prove human agency, the need for historical models transcended the divide between natural and human sciences, since even in the nonhuman world tiny variations in like objects build up so that “[t]here are . . . certain cases when variations insignificant and imperceptible in themselves are sufficient to definitively determine considerable results by way of a series of purely mechanical interactions—for example, ruptures in equilibria.”Footnote ³⁹ Because tiny variations at the level of individual objects accumulate and are magnified in complex systems, the “contingency” of Boutroux's title is not just in the laws that we use to describe nature but, to answer his opening question whether there might be a “certain degree of irreducible contingency in the world,” is a part of nature itself.Footnote ⁴⁰ Systems do not seem chaotically undetermined, but, at heart, they are indeterminate.

The first general implication of Boutroux's proposition that indeterminacy could reside in the natural world was an attack on causality. Boutroux dismissed causality as flawed on the grounds that cause and effect are based on categories, the minute imperfections of which could potentially result in an effect “failing” to follow from a cause. This means, Boutroux claimed, that causes could only definitively be attributed after the fact, so that “the cause rightly stated is only a cause if it engenders an effect,” the implication being that the material universe is not perfectly predictable.Footnote ⁴¹ Unlike the spiritualists, who searched for causality beyond the observable realm of phenomena, or the positivists, who held, in Taine's words, that “the cause of facts is in the facts themselves,” Boutroux rejected the notion of strict causality entirely, replacing it with what resembles, albeit only in retrospect, an early intimation of a probabilistic universe.Footnote ⁴²

The unpredictability of physical systems opened up a space of undetermined action and, by implication, free conscience, proof of which was the goal of Boutroux's study. Before returning to what Boutroux thought his book was about, an aside is in order. The passage on ruptures in equilibria cited above continues with an example that brings out a similarity between Boutroux's notion of contingency and later conceptions of chaos theory. The passage in its entirety reads:

It is a small step from Boutroux's bird to the famous butterfly who, in flapping her wings, causes a tornado halfway around the globe. The resemblance of Boutroux's thought to modern concepts of the indeterminacy of complex systems is neither insignificant nor fortuitous. Boutroux was tied by marriage to Henri Poincaré, who was not only one of France's greatest mathematicians, but also the father of chaos theory. Boutroux and Poincaré were in close contact in the years after Boutroux finished his thesis, forming a circle that also included Boutroux's wife, Aline Poincaré, mathematicians Jules and Paul Tannery, and astronomer Benjamin Baillaud.Footnote ⁴⁴ Poincaré read and was quite possibly influenced by Boutroux's thesis, which appeared several years before his own dissertation. Boutroux's thesis lacked scientific rigor, but his thinking was creative and his legacy fecund. The chaotic indeterminacy that Poincaré described for complex systems, Boutroux had already adumbrated for inanimate objects and had seen as central to understanding intellectual, cultural, and social systems.

Whatever small impact Boutroux may have had on mathematics, he thought that he was proving that free will and religious belief could coexist with modern physics and chemistry. The theory of a contingent physical universe allowed for radical freedom, since the products of human artifice stretch scientific categories to their limit by including so many potential imperfections that positivist attempts to study the mind as a mechanical system were bound to produce results of very limited explanatory value. Humans, residing atop the layers of complex interactions that make up the universe, are the least generalizable, least predictable, of creatures. Eclectic spiritualists had argued that mechanistic human sciences would fail because humans are radically different from the rest of the universe. Boutroux showed that even if humans are an extension and amplification of the contingent universe, positivism falls flat.Footnote ⁴⁵ A mechanistic science of man might reflect the play of patterns on the surface of society, but it would never plumb the depths, and if a science of psychology were possible, according to Boutroux, the phenomena it studied would be not “absolutely determined, but would harbor, under the uniform succession visible to the observer, a radical contingency.”Footnote ⁴⁶

Having established to his satisfaction that freedom was a fundamental condition of the universe, Boutroux faced in his conclusion the looming relativist implications of his argument. If the universe is the product of probable, but not certain, relations between objects and systems, do we not owe this universe, one in which we can live, act, and think, to random chance? Boutroux was adamant that we do not. Appealing to the idea of perfection as a kind of telos, both freely and necessarily chosen, Boutroux argued that our ideas evolve toward God. This line of thought emerged while he was writing his dissertation, when Boutroux had clearly struggled with the need to find a middle ground between atheism and religion not only in his scholarly work but in his professional life as well. His intellectual struggle was magnified by his post as a lycée teacher working for a secular state, teaching students who, he griped, lacked the refinement to grasp nuanced understandings of religion. Responding to Boutroux, who had recently complained of “nervous overexcitement” and who was possessed of a delicate constitution, Lachelier wrote,

If Boutroux was working out the conflicting demands of his secular profession and his religious beliefs in the years of teaching in Caen, he produced a provisional answer to his personal dilemma in De la Contingence. His response came in the form of a religious teleology in which the idea of “perfection” (applied to the development both of organisms and of systems) implied a goal-oriented evolution of the universe. The meaning of this universal teleology, for Boutroux, was determined by a God who was “not only the creator of the world, but also the providence who watches over the details as well as the ensemble.”Footnote ⁴⁸ The presence of such a God, according to Boutroux, explains the “contingency that is displayed by the hierarchy of forms and of general laws of the world.”Footnote ⁴⁹ Rather than devolve into caprice, our freedom is directed toward God by a new kind of causality that Boutroux introduces in his conclusion, by our drive toward “ends considered necessary, that is, good or worthy of being pursued,” a necessity that is contingent on our ability to freely and rationally choose our goals.Footnote ⁵⁰ Boutroux's conclusion, that “moral life, when practiced, appears more and more clearly . . . as the effort to be free in order to realize an end which, in itself, absolutely deserves to be realized . . . God is this end,” suggests that “perfection” is the only goal commensurate with our freedom, that to be free we must freely choose God.Footnote ⁵¹

The religiously inspired conclusion to De la Contingence brings us back to the issues raised at the beginning of this article. De la Contingence responded to the tripartite philosophical struggle over free will, but it was also a Christian argument. As one of the few modern scholars to address Boutroux's work put it, “the philosophy of contingency and creativity was the response of a believer to what he deemed the excesses of an often avowedly atheistic, materialist reductionism,” concluding from this that “Boutroux's philosophy had explicit links to religious belief.”Footnote ⁵² Were Boutroux's work only a rare case of a fervent believer accepting science, it would be enough to merit study, but both Boutroux's republican politics and the direction in which his argument was taken by his students show the extent to which not just Catholics as individuals, but the specifically Catholic concern with establishing free will to justify religious and ethical teachings, could find a home in, and could even influence, republican thought. Only by tracing Boutroux's central insight on its path through the generation that followed him can we fully appreciate the extent to which Boutroux's Catholic concern with free will developed into a pillar of French secular republicanism.

Boutroux had argued to his apparent satisfaction that modern science left room for religion, ethics, and freedom of action. The fact that in doing so he had unsettled the structure of debate over the question of free will and had brought the universally accepted claim that the material universe functioned purely mechanically into question was, to him, secondary. In retrospect, however, it would be Boutroux's willingness to accommodate his philosophy to the science and mathematics of his day that gave his work lasting import. By breaking the intellectual impasse that ensued from the decline of Cousin's orthodoxy, one in which philosophers and scientists seemed capable only of talking past one another, Boutroux guided philosophers who would base their work on scientific study, breaking ground for the philosophy of science that came to dominate the French academy in the early twentieth century under Boutroux's intellectual decedents. He also cleared a moderate political space for philosophers to reject positivism without rejecting the republicanism with which it was associated. By developing Boutroux's insight about the limited predictability of the physical world while jettisoning the religious conclusions of Boutroux's project, a coterie of his students found a way to respond to fin de siècle challenges of antirationalism and antirepublicanism.

In his native discipline, Boutroux's impact was widely felt, and his work exercised a pervasive influence on the generation who came of age around 1890 and who would go on to redraw the disciplinary map with their developments in anthropology, sociology, and psychology at the turn of the century; however, the politicization of Boutroux's defense of free will took a full twenty years to develop, and was itself contingent upon the successful establishment of the Third Republic.Footnote ⁵³ When Boutroux had first attacked positivism in the 1870s, the political implications of his work, if any, were distinctly moderate. He was defending religion, though he had also shown evidence of antiauthoritarian tendencies during his student years, when he was a signatory to a letter congratulating Senator Charles Sainte-Beuve on his stand in favor of intellectual freedom.Footnote ⁵⁴ In criticizing positivism he was undermining the philosophy of France's republican elites, but there is no evidence that he did so for political reasons. In basing his argument on what he understood to be sound scientific claims, he was rejecting the tactics of his eclectic spiritualist mentors, but he did so as much out of expediency as out of principle. By the end of the nineteenth century, however, the historicist and indeterminate epistemology pioneered by Boutroux would acquire a distinctly republican hue. This could happen because positivist philosophy was closely enough associated with the republican government that it was bound to suffer as the heroic stage of the republic passed, and the luster wore off of democratic politics. By the late nineteenth century this process was well under way, and it was increasingly obvious that secularism and science had not made men moral. The Panama scandal, in which scores of secular, republican deputies and journalists had taken bribes to hide the failures of the bankrupt Panama Canal Corporation, was a key moment. The scandal gave birth to a prototype of the new right, a coalition of anti-Semites and antirepublicans who excoriated republican politicians, but it also alienated a generation of liberal students from the “amorality” of the political elite. Even observers sympathetic to the republic had trouble arguing that a secular government was tantamount to a moral government.Footnote ⁵⁵ Boutroux's philosophy, with its overt ethical component and its acceptance of the now widely agreed-upon legitimacy of scientific knowledge, could be and was articulated as an effort to save the republic from the amorality of its founding philosophy of positivism. The recasting of De la Contingence as a republican text could only happen, however, once the overtly Catholic argument for perfection was written out. The central concern of Boutroux's thesis retained its force nevertheless, as the possibility of free and rational action was a central tenet of all of Boutroux's students’ writings. Their use of De la Contingence was a negotiation in which overt religious elements were dropped from a philosophy that remained profoundly ethically oriented. On a political level, then, Boutroux's philosophy students of the 1880s saw his idealism as a way to inject ethical demands back into a political culture that they viewed as corrupt and morally suspect, while they essentially ignored Boutroux's appeal to a Christian deity.Footnote ⁵⁶

* * *

Among those of Boutroux's students who stayed within traditional philosophy, Léon Brunschvicg hewed most tightly to the legacy of De la Contingence. Brunschvicg echoed Boutroux's emphasis on the necessity of following, rather than fighting, science. He also went further than his mentor in reevaluating their mutual Kantian heritage in light of recent developments in mathematics and physics. The external influences on Brunschvicg's thought, both his context and his personal trajectory, were, however, markedly different from those of his mentor. Rather than constructing a scientific but nondeterminist world view for religious reasons, Brunschvicg stressed the flexibility of scientific reason out of ethical, political, and secular, though still antipositivist, concerns.

Boutroux had seen his work as a response to the challenge of mechanistic determinism; Brunschvicg saw it as a historicist epistemology that could account for changing conceptions of scientific reason without, ironically, falling into either theology or relativism. For Boutroux, a divine plan had been the guarantee against the relativism of a contingent universe; for Brunschvicg, the ceaseless progress of science itself became that guarantee. Brunschvicg accepted that knowledge was historically contingent, as Boutroux had suggested when he argued that science had to take the history of beings into account; Brunschvicg pushed historicist epistemology further by suggesting that even reason evolved over time. In place of traditional correspondence theories of truth, which held that statements were true if they mirrored the external world (and which could not be historicized unless the world was changing along with scientific thought), Brunschvicg espoused an intersubjective epistemology, arguing that whatever could be agreed upon by the community of competent scientists was, in effect, the truth.Footnote ⁵⁷ For Brunschvicg and his cohort, the political importance of the historicity of truth was not so much freedom from mechanical determination as it was the fact that science was the only path to knowledge. Neither religious dogma nor the common sense valued by the eclectic-spiritualist tradition could guide moral or political decision-making because both were static, incapable of accounting for the historical development of knowledge. Privileging science and especially scientific education as the only path to the truth functioned as a defense of the secular republic at its moment of greatest need. It also helped Brunschvicg and his colleagues define an inclusive vision of French identity that was founded on education, in opposition to the exclusionary anti-Semitism they faced from anti-Dreyfusards. Coming as it did in the context of the Dreyfus affair, this transformation of Boutroux's argument from a defense of free will into a general epistemology and philosophy of science thus had significant political implications.

We can see the transformation of Boutroux's defense of free will into a defense of science, secularism, and republicanism in Brunschvicg's early work, sketched out during and just after the Dreyfus affair. The central tenet of Brunschvicg's late nineteenth-century articles was that reason must be the driving force behind sentiment, will, and even religion.Footnote ⁵⁸ Setting up his argument not as an explicit riposte to contemporaries (though he scattered clues liberally throughout his writings to suggest that he had his own time in mind), but rather as a rebuttal of Cousin, Brunschvicg attacked the eclectic-spiritualist evaluations of common sense, philosophy, will, sentiment, and reason. Cousin had, according to Brunschvicg, given common sense priority over reason, a misstep compounded by later writers who elevated sentiment and will over rational thought.Footnote ⁵⁹ Both of these positions led to intellectual errors, which Brunschvicg linked to political errors with the claim that in a democracy

In the immediate context of 1897 and 1898, when these lines were written, it is not hard to imagine who was the secular Jewish republican Brunschvicg had in mind when he worried about flawed democratic judgment. Yet Brunschvicg was also reading the events of his day through the classic French liberal problem of political judgment, raised in the aftermath of the Revolution; the classic liberal answer, limiting suffrage to those believed to have sufficient property, education, or capacité to act rationally, had been historically disgraced under Guizot. The conservative answer—reliance on tradition rather than suffrage—fails on Brunschvicg's account because without relying on reason, we have no criteria for judging between antagonistic traditions, and so the conservative solution is no solution at all.Footnote ⁶¹ Brunschvicg's response was to place reason at the core of both religion and politics, making it the driving force behind sentiment and will. Rather than limiting political participation, Brunschvicg would, in his writings and his actions, work to use the power of the state to develop citizens’ reasoning skills.Footnote ⁶² This education made sense to him because reason, according to Brunschvicg, was not an innate capacity that preceded education but rather an evolving mode of thought that required careful study and a lifelong education in the advances of science. Brunschvicg fused the republican solution to the problem of secularization (education) to a liberal question (how to reconcile freedom and popular sovereignty). Treating religion and common sense as forces with the potential to undermine political judgment, Brunschvicg laid out an argument for their mutual education at the school of reason, particularly necessary because “in the present day, when intellectual anarchy visibly decomposes and dissolves society, the work of the philosopher must be to connect moral conscience to intellectual thought, and, no less necessary for the union and progress of humanity, once more to educate common sense.”Footnote ⁶³ Arguing that all facets of social, ethical, and political calculation were compatible with reason was nothing novel, indeed it was merely recycled Kant. Brunschvicg's addition to this Kantian vision of ethics and politics was to adapt Boutroux's description of a contingent universe to address the question of what reason was. The 1890s was an era in which reason itself was under attack, on one side from conservatives, represented in France by the likes of Maurice Barrès and Ferdinand Brunetière, and on the other side from the advances in and, one might argue, a crisis of, science and mathematics that had achieved widespread philosophical recognition in the years since Boutroux's thesis. The discovery of X-rays, and with them the possibility of unmediated action over a vacuum, and especially the recognition of non-Euclidean geometries, seemingly threatened the foundations of logic and rational thought. In part due to the work of Henri Poincaré, Lobachevsky's geometry of curved space had entered the philosophical consciousness and was a growing thorn in the side of Kantian scholarship.Footnote ⁶⁴ Where Boutroux had assumed reason to be unproblematic, a constant of human nature linked to a divine plan for the universe, Brunschvicg had to explain why reason was a reliable guide once Euclid's propositions, held for 2,500 years as the paradigm of rationally derived knowledge, had proven limited in their application. He did so by making science his epistemological foundation, but he used the science of a contingent world developed by Boutroux.Footnote ⁶⁵ By accepting a probabilistic universe, Brunschvicg could describe scientific truth as approximate and reason as “flexible” (assoupli), giving it the space to grow and accommodate new discoveries in mathematics and physics.Footnote ⁶⁶

Brunschvicg's argument, that reason evolved at the pace of scientific discovery, gracefully resolved the challenge posed by scientific advances while providing a rationale for the primacy of science over common sense and religion. His reasons for doing so were at least partially politically motivated. Brunschvicg's writings of the late 1890s explicitly linked the common sense he despised in eclectic philosophy to the failure of political judgment in democratic societies, leaving little doubt that the sudden explosions of irrational anti-Semitism and antirepublicanism were on his mind.Footnote ⁶⁷ In a three-step process Brunschvicg rejected the commonsense philosophies that were complicit in the popular support of the anti-Dreyfusard movement, argued that reason was the sole legitimate foundation of political and ethical decision-making, and redefined reason (in his historical work on mathematics and physics) as a tool that evolves with the development of scientific knowledge. The implication of all of Brunschvicg's turn-of-the-century work was that scientific education was essential for good citizenship.Footnote ⁶⁸Pace early positivists, science alone could not generate ethical maxims. Only science, however, could guide the rational calculation and critical spirit necessary to act in a morally responsible way. Brunschvicg's vision of the scientific foundation of truth was so compelling that, by the early twentieth century, the epistemological tradition that he had inaugurated was widely regarded as the dominant philosophical school in France, and épistémologie was synonymous with the philosophy of science, indicative of the extent to which science had become equated with knowledge in general.Footnote ⁶⁹ Though it was Brunschvicg who provided a compelling link between the nonfoundationalist epistemology of an indeterminate universe and our moral and political world, and it was his student Bachelard who would achieve the greatest understanding of scientific method, the tradition of épistémologie would have lacked one of its most important inspirations without Boutroux's religiously motivated escape from the cul-de-sac of positivist and spiritualist debates.

It would be easy to mischaracterize Boutroux's work as Catholic apologist literature of only antiquarian interest were it not for the impact that his thesis had on later generations. Boutroux's own moderate republican commitments, as well as the political use to which his indeterminate universe was put by Brunschvicg, gives the lie to the absolute distinction that has been made between Catholic and republican thought of the nineteenth century. The historical account of science and reason, developed to leave a place for free will and religion, became the very foundation of knowledge, education, and Frenchness for Boutroux's students. The probabilistic universe, approached by historical, as well as by analytical methods, traveled through Brunschvicg's “flexible” reason and reached its acme in Bachelard's philosophy of science in the 1930s and 1940s. At the far end of the épistémologie tradition, one can still hear Boutroux's argument resonating in Michel Foucault's radical historicization of the sciences in The Order of Things. Though Foucault relied heavily on other traditions, his links to Boutroux and Brunschvicg's philosophy of science, via Bachelard, are more than genealogical; they display a distinct intellectual continuity.

The final irony of Boutroux's radical innovation in the philosophical approach to the sciences is the extent to which he failed to recognize the implications of his own work. In the early twentieth century, Boutroux began to sound increasingly like Brunschvicg when he described reason as a “living faculty” and his conception of philosophy as the “contingent and autonomous development of reason, reflecting on science and on life, which [conception] is perfectly compatible with the authority and the existence of science.”Footnote ⁷⁰ While he had begun to echo Brunschvicg's concern with the development of reason itself, a nonissue in De la Contingence, he was also backpedaling on his erstwhile claims about the natural world, mocking the idea that there could be free will in mathematics or physics.Footnote ⁷¹ Whether Boutroux, like his brother-in-law Poincaré, was worried by the relativist implications in his work, or he had truly been converted to Brunschvicg's position, is impossible to say.Footnote ⁷² Whatever the case, Boutroux's early writings had planted the seeds of a fertile tradition of nonfoundationalist philosophies of science in France, one that echoed developments outside France even while it developed through Brunschvicg and Bachelard into the épistémologie that defined French philosophy of science for the first half of the twentieth century.